spyre2

Build interactive data web apps in pure Python — no HTML, CSS, or JavaScript required.

Spyre2 is a ground-up rewrite of spyre using a modern stack: FastAPI, Alpine.js, and native support for matplotlib, Plotly, and Altair.

---

Install

pip install spyre2
pip install spyre2[matplotlib]   # + matplotlib
pip install spyre2[plotly]        # + plotly
pip install spyre2[all]           # everything

---

Quickstart

import numpy as np
import matplotlib.pyplot as plt
import spyre2

class SineApp(spyre2.App):
    title = "Sine Wave"

    inputs = [
        spyre2.Slider("frequency", label="Frequency", min=1, max=20, default=5),
        spyre2.Dropdown("color", label="Color",
                        options=["steelblue", "crimson", "seagreen"]),
    ]

    outputs = [
        spyre2.Plot("sine_plot"),
    ]

    def sine_plot(self, frequency, color):
        fig, ax = plt.subplots(figsize=(8, 4))
        x = np.linspace(0, 2 * np.pi, 500)
        ax.plot(x, np.sin(frequency * x), color=color)
        return fig

SineApp().launch()

Open http://127.0.0.1:8000.

---

Inputs

Class	Description
Slider(id, min, max, step, default)	Numeric range slider
Dropdown(id, options, default)	Select dropdown
RadioButtons(id, options, default)	Radio button group
CheckboxGroup(id, options, default)	Multi-select checkboxes
TextInput(id, default, placeholder)	Free-text input
FileUpload(id, accept)	File picker

All inputs accept a label keyword argument. If omitted, the label is inferred from the id.

---

Outputs

Class	Handler return type
Plot(id)	matplotlib.figure.Figure or plotly.Figure or altair.Chart
Table(id)	pandas.DataFrame
HTML(id)	str (HTML)
Download(id)	(filename: str, content: str | bytes)

The handler method name must match the output id:

outputs = [spyre2.Plot("my_plot")]

def my_plot(self, **kwargs):   # method name = output id
    ...
    return fig

---

Layouts

Sidebar (default)

Controls on the left, outputs stacked on the right. No configuration needed.

Grid

from spyre2 import Layout

class MyApp(spyre2.App):
    layout = Layout.grid([
        ["controls",  "controls"  ],
        ["plot_a",    "plot_b"    ],
        ["big_table", "big_table" ],
    ])

Repeat an ID across adjacent cells to span columns. "controls" is a reserved name for the input panel.

Tabs

layout = Layout.tabs({
    "Overview": ["trend_plot"],
    "Detail":   ["data_table", "bar_chart"],
})

---

Multiple chart libraries

The chart library is detected automatically from the return type — no configuration needed.

# matplotlib
def my_plot(self, x):
    fig, ax = plt.subplots()
    ax.plot(...)
    return fig              # → rendered as SVG

# Plotly
import plotly.express as px
def my_plot(self, x):
    return px.scatter(df, x="a", y="b")   # → rendered via Plotly.js

# Altair
import altair as alt
def my_plot(self, x):
    return alt.Chart(df).mark_line()...   # → rendered via Vega-Embed

---

Jupyter notebooks

app.launch() detects when it's running inside a Jupyter kernel and automatically starts the server in a background thread and displays an inline IFrame.

SineApp().launch(port=8765)   # displays inline in the notebook

---

Migrating from spyre v1

Use the included CLI tool to mechanically convert a v1 app:

spyre2-migrate myapp.py              # preview conversion
spyre2-migrate myapp.py -o new.py    # write to new file
spyre2-migrate myapp.py --in-place   # overwrite (creates .bak backup)

The tool converts imports, inputs/outputs list-of-dicts, and renames getPlot/getTable etc., leaving # TODO comments where manual cleanup is needed.

What changes:

spyre v1	spyre2
from spyre import server	import spyre2
class App(server.App)	class App(spyre2.App)
inputs = [{"type": "slider", "key": "x", ...}]	inputs = [spyre2.Slider("x", ...)]
def getPlot(self, params): x = params["freq"]	def my_plot(self, freq):
CherryPy on port 9093	uvicorn on port 8000

For apps that can't be fully migrated yet, spyre2.compat.App accepts the old dict-based syntax (deprecated, will be removed in a future release).

---

Examples

File	Demonstrates
examples/sine_example.py	Sidebar layout, matplotlib, slider + dropdown
examples/grid_example.py	Grid layout, matplotlib, multiple outputs + table
examples/plotly_example.py	Tabs layout, Plotly, scatter + histogram